Studies are proposed to advance understanding of the organization of human memory and the nature of memory disorders. The work involves six studies, which are organized around three topics that have been prominent in recent discussions of memory and memory impairment: A) Recent memory and remote memory; B) Recognition memory and the human hippocampus; C) Working memory and the medial temporal lobe. A salient aspect of the proposed work is the opportunity to continue study of our population of memory-impaired patients with bilateral lesions limited to the hippocampus or with larger medial temporal lobe lesions. Many of these patients are veterans. Our population of amnesic study patients, including the veterans, is one of the very few such populations available anywhere and is the best characterized in terms of detailed, quantitative neuroanatomy. A second and complementary aspect of the proposed work is neuroimaging studies of learning and memory in healthy individuals using functional magnetic resonance imaging (fMRI). The following six studies are proposed: A) Recent memory and remote memory (an fMRI study of healthy volunteers). This is a prospective study of recent and remote memory, testing memory in the scanner for material that was learned 1 hr to 1 month earlier. We will also determine how brain activity in response to the new items that are presented during scanning is related to how novel these items are judged to be during scanning and how well these items are remembered in a later memory test. In this way, we will assess brain activity both as it relates to the age of a memory (1hr to 1 month) and also as it relates to the formation of new memories. Ba) Recognition memory and source memory (an fMRI study of healthy volunteers). We will use an improved and novel design asking whether activity in the hippocampus and perirhinal cortex during learning is differentially predictive of subsequent recognition memory and the subsequent ability to remember the context in which material was learned. This study addresses a fundamental issue: how sharp is the division of labor within medial temporal lobe structures? Bb) Remember (R) and Know (K) responses after hippocampal damage. We will ask whether the hippocampus selectively supports recollection (i.e., memory for the context in which learning occurred), or whether the hippocampus has a broader role in memory that includes the ability to detect simple familiarity. Bc) Recognition memory for faces. We will ask whether the capacity for face learning and face recognition is organized differently than the capacity to learn and recognize other kinds of material (e.g., specifically words and scenes). Is face memory hippocampus-dependent? Ca) Working memory for object-place associations. We will use a novel change-detection task to determine the working memory capacity for object-place associations and will ask whether working memory for this domain of material is independent of the medial temporal lobe. Cb) Relational working memory at short lag times. We will ask whether working memory is intact or impaired after medial temporal lobe damage as measured by tasks that assess relational memory for locations within complex scenes. These two studies (Ca and Cb) reassess the classical distinction between short-term and long-term memory. The issue hinges on the difficulty in determining which process (short-term memory or long-term memory) is supporting task performance in any particular case. I have devised a method that can make this determination.